Integrated Learning Particle Swarm Optimizer for global optimization

This study proposes a novel Integrated Learning Particle Swarm Optimizer (ILPSO), for optimizing complex multimodal functions. The algorithm modifies the learning strategy of basic PSO to enhance the convergence and quality of solution. The ILPSO approach finds the diverged particles and accelerates them towards optimal solution. This novel study also introduces the particle’s updating strategy based on hyperspherical coordinates system. This is especially helpful in handling evenly distributed multiple minima. The proposed technique is integrated with comprehensive learning strategy to explore the solution effectively. The performance comparison is carried out against different high quality PSO variants on the set of standard benchmark functions with and without coordinate rotation and with asymmetric initialization. Proposed ILPSO algorithm is efficient in terms of convergence rate, solution accuracy, standard deviation, and computation time compared with other PSO variants. Friedman non-parametric statistical test followed by Dunn post analysis results indicate that the proposed ILPSO algorithm is an effective technique to optimize complex multimodal functions of higher dimension.